Large-scale molecular analysis is central to understanding a wide range of biological phenomena related to states of health and disease both in humans and in a host of economically important plants and animals, e.g. Collins et al (2003), Nature, 422: 835-847; Hirschhorn et al (2005), Nature Reviews Genetics, 6: 95-108; National Cancer Institute, Report of Working Group on Biomedical Technology, “Recommendation for a Human Cancer Genome Project,” (February, 2005). Miniaturization has proved to be extremely important for increasing the scale and reducing the costs of such analyses, and an important route to miniaturization has been the use of microarrays of probes or analytes. Such arrays play a key role in most currently available, or emerging, large-scale genetic analysis and proteomic techniques, including those for single nucleotide polymorphism detection, copy number assessment, nucleic acid sequencing, and the like, e.g. Kennedy et al (2003), Nature Biotechnology, 21: 1233-1237; Gunderson et al (2005), Nature Genetics, 37: 549-554; Pinkel and Albertson (2005), Nature Genetics Supplement, 37: 511S17; Leamon et al (2003), Electrophoresis, 24: 3769-3777; Shendure et al (2005), Science, 309: 1728-1732; Cowie et al (2004), Human Mutation, 24: 261-271; and the like. However, the scale of microarrays currently used in such techniques still falls short of that required to meet the goals of truly low cost analyses that would make practical such operations as personal genome sequencing, environmental sequencing to use changes in complex microbial communities as an indicator of states of health, either personal or environmental, studies that associate genomic features with complex traits, such as susceptibilities to cancer, diabetes, cardiovascular disease, and the like, e.g. Collins et al (cited above); Hirschhorn et al (cited above); Tringe et al (2005), Nature Reviews Genetics, 6: 805-814; Service (2006), Science, 311: 1544-1546.
Increasing the scale of analysis in array-based schemes for DNA sequencing is particularly challenging as the feature size of the array is decreased to molecular levels, since most schemes require not only a procedure for forming high density arrays, but also repeated cycles of complex biochemical steps that complicate the problems of array integrity, signal generation, signal detection, and the like, e.g. Metzker (2005), Genome Research, 15: 1767-1776; Shendure et al (2004), Nature Reviews Genetics, 5: 335-344; Weiss (1999), Science, 283: 1676-1683. Some approaches have employed high density arrays of unamplified target sequences, which present serious signal-to-noise challenges, when “sequencing by synthesis” chemistries have been used, e.g. Balasubramanian et al, U.S. Pat. No. 6,787,308. Other approaches have employed in situ amplification of randomly disposed target sequences, followed by application of “sequencing by synthesis” chemistries. Such approaches also have given rise to various difficulties, including (i) significant variability in the size of target sequence clusters, (ii) gradual loss of phase in extension steps carried out by polymerases, (iii) lack of sequencing cycle efficiency that inhibits read lengths, and the like, e.g. Kartalov et al, Nucleic Acids Research, 32: 2873-2879 (2004); Mitra et al, Anal. Biochem., 320: 55-65 (2003); Metzker (cited above).
In view of the above, it would be advantageous for the medical, life science, and agricultural fields if there were available molecular arrays and arraying techniques that permitted efficient and convenient analysis of large numbers of individual molecules, such as DNA fragments covering substantially an entire mammalian-sized genome, in parallel in a single analytical operation.